sradmard's profile - activity

So we are clear that rviz subscribes to both topics, but now the question is, which topic does dwa_local_planner under m

Another workaround in publishing every updates into the full costmap is to set the following parameter: always_send_f

The /costmap publishes the costmap values while the /costmap_updates returns the value of the updated area of the costma

You are right, rviz seems to subscribe to both /costmap and /costmap_updates topics. However, I cannot find any document

Hi, I am encountering the issue that my local cost map does not refresh when the robot is stationary. This would create significant problems when there are moving obstacles around the robot. You can observe the behaviour here at 0:22 Local_costmap_issue Here is how I launch my move_base node in my robot launch file:

<node pkg="move_base" type="move_base" respawn="false" name="move_base" output="screen">
<rosparam file="$(find turtlebot_vrep)/config/costmap_common_params.yaml" command="load" ns="global_costmap" />
<rosparam file="$(find turtlebot_vrep)/config/costmap_common_params.yaml" command="load" ns="local_costmap" />
<rosparam file="$(find turtlebot_vrep)/config/local_costmap_params.yaml" command="load" />
<rosparam file="$(find turtlebot_vrep)/config/global_costmap_params.yaml" command="load" />
<rosparam file="$(find turtlebot_vrep)/config/base_local_planner_params.yaml" command="load" />
<rosparam file="$(find turtlebot_vrep)/config/move_base_params.yaml" command="load" />
<remap from="cmd_vel" to="cmd_vel_mux/input/navi"/>
</node>

Relevant data are getting published on /map topic at about 16 Hz, on /scan topic at 20 Hz, on /odom topic at 20 Hz, and on /tf topic at about 300 Hz. However, the /move_base/local_costmap/costmap topic barely gets published on, and does not publish anything when the robot is stationary.

Also, here are my parameters:

controller_frequency: 20.0

TrajectoryPlannerROS:
  max_vel_x: 10.45
  min_vel_x: 0.1
  max_vel_theta: 10.0
  min_in_place_vel_theta: 0.4
  acc_lim_theta: 3.2
  acc_lim_x: 2.5
  acc_lim_y: 2.5
  holonomic_robot: true

global_costmap:
  global_frame: map
  robot_base_frame: base_link
  update_frequency: 1.0
  static_map: true
  publish_frequency: 0.5
  transform_tolerance: 0.5

local_costmap:
  global_frame: map
  robot_base_frame: base_link
  update_frequency: 5.0
  publish_frequency: 2.0
  static_map: false
  rolling_window: true
  width: 3.0
  height: 3.0
  resolution: 0.05

Move base node parameters.

shutdown_costmaps: false
controller_frequency: 5.0
controller_patience: 3.0
planner_frequency: 1.0
planner_patience: 5.0
oscillation_timeout: 10.0
oscillation_distance: 0.2

Costmap common psrsmeters:

transform_tolerance: 1
obstacle_range: 2.5
raytrace_range: 3.0
footprint: [[0.25, 0.2], [-0.25, 0.2], [-0.25, -0.2], [0.25, -0.2]]
inflation_radius: 0.35
observation_sources: laser_scan_sensor
laser_scan_sensor: {sensor_frame: Hokuyo_URG_04LX_UG01_ROS, data_type: LaserScan, topic: /scan, marking: true, clearing: true}

I get the same behaviour on the actual turtlebot as well, and I would really appreciate any help.

Finally, it appears that, the /move_base node in fact adds new info to the costmap, but it does not execute clearing act

https://youtu.be/gvx-dNJ-NVY The second question is, what do the "update_frequency", and "publish_frequency" params do

Thank you so much for your answer. There are two questions here: 1 I do not know how to check for true publishing rate o

Local Costmap of move_base Package Does not Refresh when the Robot is Stationary. Hi, I am encountering the issue that m

Local Costmap of move_base Package Does not Refresh when the Robot is Stationary. Hi, I am encountering the issue that m

Therefore, when a dynamic obstacle is recorded as an obstacle on the costmap, it takes a while for the occupancy grid to

Sorry for the late response, and thank you for your comment. What I meant by no clearing actions is that OccupancyGrid i

Local Costmap of move_base Package Does not Refresh when the Robot is Stationary. Hi, I am encountering the issue that m

How to Implement twist_mux Properly? Hi there, I am trying to provide velocity commands to control my robot, but these

How to Install rospy Locally to Create an Amazon AWS Lambda Function? Hi everyone, I am trying to use Amazon alexa to c

I found the solution to the problem and wanted to share. The source of the problem was using v-rep simulation environment to publish point cloud data. I used the publisher “simros_strmcmd_get_depth_sensor_data” in v-rep to publish PointCloud2 data to ROS. The current version of v-rep publishes the frame_id of the header of the published PointClouad2 message in ROS with a preceding “/”. This was compatible with tf, but is not compatible with tf2 standard since there was a change in tf2, that frame_ids cannot start with an “/” anymore. Therefore, there will be a problem in ROS if we try to apply tf2 to the published data of “simros_strmcmd_get_depth_sensor_data”. I managed to debug the “simros_strmcmd_get_depth_sensor_data” source file so that it would publish PointCloud2 with a frame_id that does not contain a preceding "/". More detail on the debugging is provided here: http://www.forum.coppeliarobotics.com/viewtopic.php?f=5&t=6188

I was able to fix the problem by making sure that the frame attached to the base_link has its z-axis pointing upward. You can check it through rviz by visualizing tf and selecting individual frames to visualize.

Hi,

I am using ROS indigo under UBUNTU 14.04. I am also simulating a turtlebot equipped with kinect camera in v-rep and would like to run gmapping through ROS. Since gmapping requires laserscan data, and I can only publish pointcloud data to ROS from v-rep, I need to use pointcloud_to_laserscan package. However I haven't been successful so far, and this is what I've done so far:

I can publish the simulated kinect data to ROS as sensor_msgs/PointCloud2 type on a topic called /cloud_in. This pointcloud data is being published with respect to frame_id of base_scan. I also publish some frames in my tf tree including world, base_link (the robot frame), and base_scan (the kinect frame) with the following hierarchy:

world -> base_link -> base_scan

Then, when I run my pointcloud_to_laserscan launch file, the laserscan data is supposed to be published on /scan topic. However, depending on what I set for the target_frame parameter in my launch file, I run into different issues. Here are the different cases:

1- If "target_frame" is not set in the launch file, then it will publish laserscan data on /scan topic with respect to the same frame_id of base_scan. If I echo the pointcloud data from /cloud_in topic, I see the frame_id to be /base_scan. If I also echo the /scan topic, its frame_id is also /base_scan. The problem in this setting is that /scan topic returns manly inf values, which is incorrect given my simulated environment. In addition, I do not get anything by running the following command:

$ rosparam get /pointcloud_to_laserscan/target_frame

From my understanding of similar posts on the web, this is apparently because of incorrect frame setting and transformation, so I did the following next.

2- If "target_frame" is set to base_link or world in the launch file, then nothing gets published on /scan topic. By running rqt_console, I see the following error:

Transform failure: Invalid argument "/base_scan" passed to lookupTransform argument source_frame in tf2 frame_ids cannot start with a '/' like:

Location:~/v-rep_ws/src/pointcloud_to_laserscan/src/pointcloud_to_laserscan_ nodelet.cpp:PointCloudToLaserScanNodelet::cloudCb:185

In this case the following command returns base_link or world depending on the launch file parameter.

$ rosparam get /pointcloud_to_laserscan/target_frame

3- If "target_frame" is set to /base_link or /world in the launch file, then nothing gets published on /scan topic again. By running rqt_console, I see the following error:

Transform failure: Invalid argument "/base_link" passed to lookupTransform argument target_frame in tf2 frame_ids cannot start with a '/' like:

Location:~/v-rep_ws/src/pointcloud_to_laserscan/src/pointcloud_to_laserscan_ nodelet.cpp:PointCloudToLaserScanNodelet::cloudCb:185

In this case the following command returns /base_link or /world depending on the launch file parameter.

$ rosparam get /pointcloud_to_laserscan/target_frame

I would really appreciate any hint or help on these issues.